Vol.12_No.2 - Pesticide Alternatives Lab - Michigan State University

Spring 2003 Resistant Pest Management Newsletter Vol. 12, No.2development (Georghiou & Taylor 1977) and thetheoretical potential for spraying crops with extremelyhigh doses of one or more insecticides has beendiscussed often (Roush 1989, Tabashnik & Croft1982). The "high-dose refuge" strategy is the mostwidely used and has been implemented in NorthAmerica (Alstad & Andow 1995). When an insecticidespray kills 95% of the susceptible (SS) individuals, thesurvival of RS individuals is likely to be significantlyhigher, unless the alleles governing resistance happento be phenotypically recessive (i.e, the RS and SSinsects are physiologically identical). Instead of hopingthat resistance is phenotypically recessive, the highdose approach attempts to make resistance alleles"effectively recessive" even if they are notphenotypically recessive (Gould 1998). Similarly, dosethat is insufficient to kill the insects bearing one copyof a major resistance allele renders resistancefunctionally partially dominant. Hence, the onlycommercially available approach to reduce thelikelihood of resistance development is the use of ahigh dose of a single gene, producing 25 times thetoxin concentration needed to kill susceptible insects incombination with a refuge.High concentrations of Cry1Ac in bolls oftransgenic cotton are essential for achievingfunctionally recessive inheritance of resistance (Liu etal. 2001). Further, extensive planting of transgenic cornhybrids having sub-optimal production of the toxin andresulting in only moderate effects on H. zea wouldraise concerns about the rapid evolution of resistance(Storer et al. 2001). If transgenic plants could be madeto express enough toxins to overcome all homozygousresistance alleles, the crop in question would become anon-host. The lack of a "high dose" in current Bt cottoncultivars for H. armigera and the small scaleproduction systems of cotton indicates that the "highdose/refuge" resistance management strategy is notfeasible for Bt cotton in northern China (Zhao et al.2000a). Under these circumstances, supplementalcontrol of H. armigera with insecticides is essential togrow Bt cotton for a longer period (Ru et al. 2002).Resistance in insects to Bt can be dramatically reducedthrough the genetic engineering of chloroplasts inplants. Several copies of the Bt genes could beexpressed per cell via the chloroplast genome asopposed to only two copies via the nuclear genome in adiploid cell. The Cry2Aa2 protoxin levels inchloroplast-transformed tobacco leaves are between 2to 3% of total soluble protein, and are 20-to-30-foldhigher than current commercial transgenic plants (Kotaet al. 1999). If a toxin is consistently produced by aplant at a highly toxic concentration without having anegative effect on yield, and the toxin does not affectnon-target organisms, then the constraints on high dosestrategy would be quite low.Another serious concern regarding the success ofhigh dose strategy is that the hypothesis of resistancebeing recessive does not hold in different insectspecies. Inheritance of resistance showed incompletedominance in O. nubilalis to a commercial preparationof Bt (Huang et al. 1999), and in H. virescens toCry1Ab (Sims & Stone 1991). While, Tabashnik etal.(1998) demonstrated dominant resistance to Cry1Aain a strain of P. xylostella having field-evolved Btresistance.CONTROLLED EXPRESSION of TOXINS Mono-cultivationof Bt transgenic crops is likely to select intensely forresistance because pests will be exposed to Bt evenwhen they are not causing economic damage (Mallet &Porter 1992). The degree of yield reduction caused by apest population is dependent on its density, as well ason when and where insects feed on the plants.Expression of toxin coding genes could be limited tovulnerable plant parts, and at times when toxicity isneeded most. If a pest causes no damage when it feedson mature leaves, but causes severe stunting when itfeeds on buds and developing leaves, then toxinproduction only in buds would be useful. Having Btexpressed in plants so that the insect population issubjected to selection pressure for particular periods oftime (e.g., through an inducible promoter) or inparticular plant parts (e.g., through tissue-specificpromoters) may provide larger refuges for susceptiblealleles both within the field and within a region whileat the same time minimizing the crop loss (Roush1997b). This can be achieved by using gene constructshaving a tissue specific promoter.In P. xylostella, resistance to Bt declined whenexposure to insecticide ceased (mean R = -0.19). Infour other pests (H. virescens, L. decemlineata, Muscadomestica and P. interpunctella), resistance to Btdeclined slowly or not at all (mean R = -0.02) in theabsence of exposure to Bt (Tabashnik et al. 1994 ).Similar loss of resistance in O. nubilalis was observedin the absence of selection pressure (Bolin et al. 1999).This can be exploited for formulating resistancemanagement strategies by enforcing completerestriction on cultivation of certain Bt cultivars for aspecified period.Solutions to resistance management involvecomplex strategies. The track record of resistancemanagement for chemical pesticides is notencouraging. The wisdom gained from previouspesticide failures should provide impetus for theproactive development and implementation ofmanagement strategies for transgenic crops. Keepingthis in view, Cohen (2000) made four practicalrecommendations for promoting the sustainable use ofBt crops, based on existing knowledge of the principlesof resistance management:22